I’m a senior software engineer working in the heart of our digital identity and trust team. We’re a close-knit team of specialists and we’d been used to working physically close to one another when we were office based. Now, of course, we’re all working from home, which has its drawbacks but overall it’s working remarkably well.
Since mid-March we’ve made a really good transition to home-working, using the engineering tools and techniques available to us to drive forward our digital transformation.
Why we’re developing automation
We’re optimising our investment in automation to allow us to free colleagues from repetitive manual processing. Our team is focused on transitioning all of our digital services from being essentially 1-way, information only flows in one direction, from us to our users to 2-way allowing our users to interact and respond. We’re using a pool of technologies known as the Dynamic Trust Hub to do this. Users usually interact with DWP by filling in an online form which is then processed manually. Our goal is to shift some of the burden of this away from us and back to the user, allowing them to create a verified and trusted online identity which can be re-used.
Testing and automation
In our development and testing work we’re also big believers in automation. Not only because it saves time in the long run, but also because it vastly reduces the risk of errors which can arise with manual processes, configuration and testing.
With a product as complex as the Dynamic Trust Hub, we brought in automation at the very start. Scripting our environments so that they could be set up with minimal intervention. Building a full suite of automated tests for the cases and user journeys, and automating code quality checks to ensure that coding errors or mistakes were not inadvertently introduced.
Using AWS and innovative tech for fast deployment
Cloud services and tooling allow us to deploy software more quickly and put the user at the centre of customer-led journeys. Over the past few years there has been a big move to cloud, chiefly utilising Amazon Web Services (AWS). For an organisation like ours, the big advantage of using cloud-based services is that we can rapidly scale up as needed, while also being able to take advantage of some of the cost savings associated with hosting and managing our own data centres.
AWS offers an excellent range of services and the security and resilience of their platform also makes it easy for us to deploy new services as we need them. For example, to set up a server for performance testing which can then be brought down when it’s no longer needed.
Developing secure identity software
To help achieve our goals of security and reliability for users, we’re using one of the leading identity and access management off-the-shelf packages, ForgeRock, to drive much of the functionality. ForgeRock includes BBC iPlayer as one of its big success stories, and I was part of the team of developers which evaluated the product before it was selected to be used by us.
My current work with ForgeRock is really varied and interesting, and uses a good range of technology. ForgeRock includes a complete RESTful API layer, which means we’re working on 2 primary interfaces, a standard GDS Node app/front end, and a 3rd party voice-activated system which will allow users to interact with DWP via ForgeRock services over the phone. It’s my first opportunity to work with voice-activated systems and speech recognition, so this has been especially interesting.
Working with voice recognition
We’re working with a team of technicians to refine the speech recognition interface between the user and the ForgeRock layer. DWP has a lot of hands-on experience of dealing with some of the intricacies and subtleties of voice recognition. We use our own algorithms to help decipher what users are trying to tell us and turning this into a string of text and a probability score. Currently a majority of our users still prefer to interact with us over the phone rather than by post or online. This means that speech recognition is likely to be an important way to make communication with the department easier until our users feel comfortable with the ‘channel shift’ from phone to online.
Developing and increasing software potential
ForgeRock software provides a lot out-of-the-box, but what makes it quite powerful is the ability to use ‘scripted nodes’ and ‘custom nodes’ to extend the functionality as you want. We’re developing one-off scripted nodes in JavaScript, and new custom nodes written in Java for reusable pieces of functionality.
ForgeRock isn’t the only software suite we’re using. Within my specific squad, we’re developing an ‘ID Store’. This is a micro-service which captures ID verification events and allows us to verify a users’ identity by querying the ID events we know about them.
There’s a huge amount going on in identity and trust, with real potential to provide transformational ways in which users can interact with us.
In the long term, there’s a lot of exciting technology we are looking at. Transaction risk is another area we’re exploring, which will allow us to use a range of data sources to consider each interaction with the Dynamic Trust Hub and respond accordingly.
With all of the new innovation in our teams it’s an exciting time to join us. Have a look at our current vacancies to apply now.
2 comments
Comment by Michelle posted on
wow, this sounds really exciting stuff
Comment by Rashmi posted on
Very informative blog!